7F Users Group – Unwavering commitment

Unwavering commitment to the industry at 20 years and counting

The 7F Users Group hosted its 20th anniversary conference last May in Houston’s Westin Galleria, attracting about 250 owner/operators to the world’s largest gathering of F-class users and the equipment and services providers that serve them.

The organization has grown dramatically since its first meeting in Baltimore, Nov 19-20, 1991, attended by 14 O&M personnel from four companies. The only 7F engine in service at that time was Virginia Electric & Power Co’s Chesterfield 7. It had been integrated with an existing steamer and new heat-recovery steam generator to configure a 1 × 1 combined cycle.

 

Fast forward to today: The OEM has shipped to customers more than 1000 F-class gas turbines since Chesterfield 7 left the shop. The large majority of these units are 7FAs.

The meticulous minutes of the first meeting focus on the gas turbine (7F history, p 34). Recent conferences have been multi-dimensional, with robust sessions on HRSGs, steam turbines, and electric generators, in addition to those on the engine’s compressor, combustion system, turbine, auxiliaries, and controls.

Another big difference between the meetings of yesteryear and today: The minutes of the first meeting don’t mention the OEM’s participation. The Houston program, which began Monday, May 9, offered two days of OEM presentations, reflecting how F-class machines have grown in complexity over the years to satisfy ever more demanding environmental regulations and owners’ goals of higher efficiency and availability/reliability.

The first meeting also did not have a vendor fair or vendor presentations. The 2011 conference had vendor fairs on three nights, one exclusive to the OEM’s products and services. More than 100 companies participated in the exhibition the first two evenings.

The 2011 technical program included (1) informal user presentations, (2) open discussions, and (3) special presentations by representatives of 15 third-party vendors. The first two are summarized on the next several pages; vendor presentations are in the second half of this report, beginning on p 16. The steering committee—expert user volunteers all—developed the program and conducted the meeting (Sidebar 1). Sheila Vashi and her colleagues at Vision-Makers organized the event. There were 14 sponsors (Sidebar 2).

Day One

It was a long opening day. Some attendees gulped a quick cup of breakfast coffee and boarded busses at 7:30 for a tour of GE Energy’s Houston shops, others were on the golf course by 8:00 (see p 33), still others participated in the special HRSG session from 8:30 to 12:30 conducted by technical experts from HRST Inc.

In the afternoon, an hour-long introductory session designed to prep first-timers (perhaps one-third of the registrants) for the user-only discussion sessions on Tuesday and Wednesday preceded a two-hour OEM session on controls.

So when the 7F faithful mustered for the Keynote at 6:30 pm, the only thing between them and refreshments was G Michael Curley, manager of GADS Services for the North American Electric Reliability Corp. His message, in brief, was more operating data are wanted from generating facilities.

Talk about a tough assignment. But Curley was equal to the task. His mission was to give owner/operators a heads-up on NERC’s proposal to collect reliability data and information on commercial generating units.

He began by reminding the users of NERC’s mission: To ensure the reliability of the bulk power system. Curley said NERC has made great progress in this regard without undue burden on stakeholders by carefully evaluating what industry data are required to achieve assigned objectives, by reviewing and using to the extent possible available information resources, and by modifying existing programs as needed to meet goals.

In June 2010, he continued, the NERC Planning Committee created a task force to review if its Generating Availability Data System (GADS) should be mandatory reporting for all generator owners on the NERC Compliance Registry—specifically, all commercial generating units 20 MVA and larger or generating facilities 75 MVA and larger. Those in the room with gray hair recalled that NERC has used GADS to collect generating-plant availability information for nearly three decades.

The task force recommended that GADS be mandatory, Curley said, but at the same time suggested minimizing the amount of data collected by focusing only on those characteristics affecting bulk power system reliability. This would reduce by about half the information requirements of “full GADS,” which many in the room were familiar with.

The GADS Task Force recommended using Section 1600 of NERC’s “Rules of Procedure” for data collection. Under this procedure, GADS data collection is not governed by NERC standards and is not under NERC enforcement.

In fact, Curley added, NERC cannot apply financial penalties for delayed or incomplete reporting—or for not reporting any of the information requested. However, those companies not providing data would be in violation of the “Rules of Procedure” under Section 215 of the Federal Power Policy Act of 2005.

Late last summer, the NERC Board of Trustees agreed to a phase-in of the commercial generating units: 50 MW and larger starting Jan 1, 2012 and 20 MW and larger starting Jan 1, 2013. Note that the final rules are based on unit nameplate capacity only, not facility generating capability as originally proposed by the trustees and cited by Curley during his presentation before the 7F users. Curley told the editors in mid December that plants already reporting to GADS would see no difference in their reporting requirements now than they did before mandatory GADS.

Day Two: Compressor section

The Houston technical program held with tradition. Informal presentations by users on compressor issues headlined the opening session, followed by open discussion (Sidebar 3). The first speaker, who has responsibility for eight 7FAs (7241s), told the group that eddy-current inspection of one unit, which had accumulated 942 starts and 14,000 hours of service since commercial operation began eight years ago, revealed 14 cracked S0 vanes. Two other GTs each had two vanes with cracks; another, one.

Eddy-current inspections are conducted semi-annually. Annually, dental molds are taken and each engine is borescoped to assess the condition of the compressor and to check for such things as vane platform-stepping and shim migration.
All cracked vanes described above were found in the lower half of their respective units, most between 5 and 7 o’clock; worst crack, 2 in. long, was found in the 6 o’clock airfoil at mid span. Rotor removal was required to extract the lower-half vane rows.

The gas turbines are located in a high-humidity environment and the vane rows were “welded” in place by rust. Online water washing and fogging were terminated in 2006. Analyses conducted by two independent engineering firms pointed to corrosion and loss of airfoil damping as the root cause of cracking. A user in the audience suggested there were other factors as well—perhaps thinning of the trailing edge.

S0, S1, and S4 vanes were replaced in the four affected gas turbines, S2 and S3 in three of the four. Plans are in place to remove OEM vanes in the four engines without vane distress during their next hot-gas-path inspections.

Replacement hardware was provided by a third-party supplier; the OEM never submitted a bid. Field service was performed by third-party contract personnel.

Inspection experience. The editors recalled a conversation with Rod Shidler of Florida-based Advanced Turbine Support Inc a month earlier when he said ATS inspection teams had found S0 cracking in at least three-dozen units during the last year, including one engine with fewer than 1000 fired hours and only 151 starts. He believes that perhaps no 7FA can be considered immune to S0 cracking. During a recent visit to a four-unit plant, he found three engines with cracked S0 vanes.

The consensus view of participants in the GE Roundtable at the CTOTF’s Spring Turbine Forum was that S0 cracking is an emerging fleet issue and that every plant should schedule a borescope inspection when convenient. At that session, it was said that eddy-current inspection was successful in identifying S0 cracks; also that indications won’t necessarily bleed dye during the alternative penetrant inspection. Attendees were urged to consult the latest version of Technical Information Letter (TIL) 1509.

Webinar. To be sure 7F owner/operators were up to speed on what may be the most important OEM advisories for plants with this engine, ATS developed PowerPoint presentations on TILs 1509, 1638, and 1795 and communicated the material via a December 14 webinar. Mike Hoogsteden and Dustin Irlbeck of ATS presented the material and CCJ ONscreen’s Scott Schwieger coordinated and produced the event. Turn to p 36 for an article summarizing the material disseminated.

Filter glop fouls compressor. The background facts: A 2 × 1 combined cycle equipped with three-year-old horizontal conical/cylindrical inlet filters and no evaporative cooling system. Compressor cleaning was restricted to a semi-annual offline water wash. One day, GT 1 was found producing 27 MW less than its sister unit. Inlet filters were inspected and recorded as “wet.” Pressure drop across the Unit 1 filters was double that across Unit 2’s. Intermittent increases in engine vibration also were noted.

The engine was removed from service and inspected. Airfoils at the front end of the compressor were fouled with “sludge,” those in the back of the unit covered with a white powder. There was so much sludge on the compressor blades, the user presenter said, you could grab handfuls. Standing water was identified just ahead of the air filters and just ahead of the trash screens.

The wet filters were sent out for inspection and evaluation. They were found loaded with spring pollen. Cooling-tower drift was identified as the source of the moisture. Drift never affected the sister unit simply because of Unit 2’s more favorable location.

Two online water washes cleaned the sludge off the airfoils. No blade or vane damage was found. The vibration spikes were thought caused by clumps of sludge being released from upstream airfoils and impacting downstream rotor blades.

The fix: The conventional conical/cylindrical filters were replaced with triple-wrap HEPA (H12) filters and drift eliminators were installed ahead of the filter bank. It was too early to tell how effective the solution described is; the new filters had only been in service for a couple of weeks at the time of the meeting.

This case history introduced some serious discussion on the service life of filters, the viability of HEPA filters as an alternative to conventional ones, etc. A participant in the conversation said he has filters evaluated annually for effectiveness. At one of his plants, the results dictate changing filters every two to three years; at another, he believes he’ll get to the second major (12 years) on the original filters. Note that prefilters are used at the second location and are changed quarterly.

The next presentation discussed R0 issues: biscuit rotation, blade staking, and erosion attributed to the OEM’s SPRITS™ power augmentation system. The speaker representing a 2 × 1 combined cycle reported that the R0 biscuit mod had been implemented and that new R0 enhanced blades had been installed in March 2010. The latter had operated for 2000 hours and nearly 150 starts before inspection revealed a gap on one blade.

Closer inspection revealed that one blade had walked into the inlet guide vanes (IGVs) and started to machine itself. The contact was between the blade and the rub-ring area of the bellmouth. The biscuit had rotated 180 deg and allowed the blade to walk. Attendees were made aware that the recently released (Apr 25, 2011) TIL 1797 addresses the issue and advised to inspect at the next opportunity. In this case, experts believed staking was not sufficiently robust to deform the metal and lock in the biscuit.

The OEM was said to have told users that new R0 blades can tolerate erosion pitting because of the high compressive stress imparted by laser shot peening. Reportedly, as long as the depth of the pits does not exceed the depth of the peen—about 50 mils—allowable stresses will not be exceeded. Experience suggests that Sprits erosion will be about 10 mils after about 600 hours of fogging.

Don’t forget the inlet screens. The final user presenter on compressors said that two 7FAs at his plant were found to have R0 impact damage. Small pieces of stainless steel wire were found in the inlet. Examination of the inlet screens revealed they were the source of the loose pieces of wire. The incident added seven days to a CI. The speaker urged his colleagues to make screens a key inspection point and not to hesitate replacing them when fraying begins. He also mentioned a redesigned screen was available, one less prone to causing domestic object damage.

Combustion section

An offshore owner/operator with four 1 × 1 7FA+e-powered combined cycles reported on a fractured diffusion-air pipe issue associated with a DLN2.6 combustor during the 2010 meeting. The underlying cause of the failure, he believed, was poor welding and high stress attributed to vibration. Further research and analysis were required.

The user returned this year to finish the story. The pipe, which had been repaired shortly before the 2010 meeting, broke again three months later. Comparison photos of the broken metal showed different failure patterns, leading the user to conclude that the cause of the second event was not the same as that for the first.

Vibration levels and the stress on the pipe were measured under varying operating conditions. Strain amplitude was found proportional to power output. The investigators concluded that high vibration was the primary cause in the first event and high-cycle fatigue secondary. For the second event, the primary cause was defective welding, vibration secondary.

Several more end covers were found with diffusion-air pipes having substandard welds and they were repaired. The OEM was challenged to investigate its design, manufacturing, and quality practices. The result: A welding process issue was identified and some quality-control procedures found inadequate. Related processes and procedures were under review at the time of the 2011 meeting and will be modified as necessary.

The user said engine vibration was reduced by use of balance shots. The unit, which had been operating at 3 mils is now running smoothly at less than 1 mil.

Automated combustion tuning. You can put a point on the electric power industry’s technology timeline at 2011 for commercial acceptance by owner/operators of automated gas-turbine combustor tuning systems. That was the conclusion of the editors based on presentations (user and vendor) and discussions at the 7F Users Group meeting.

The dry low-NOx (DLN) combustion systems that helped gas turbines become the preferred generation alternative made users aware of “combustion dynamics.” Recall that DLN combustion systems are prone to flame instability under certain operating conditions and the resulting pressure pulsations (dynamics) can be strong enough to damage both combustors and downstream hot-gas-path components.

To mitigate the destructive effects of these pulsations, OEMs participated in the development of combustion dynamics monitoring systems. The CDMS warns operators of impending instability, enabling them to make the necessary adjustments to maintain stable combustion.

The early CDMSs, which debuted only a few years ago, were “manual” systems. The ratcheting downward of emissions limits and plant staffs helped put the development of automated systems on the fast track. A user presentation updated the group on the status of tuning systems, which have automated the functions of the CDMS and integrated them into a controls package that assures optimal operation of the gas turbine on a continuous basis.

The user-presenter, who is the control system troubleshooter for a major F-class fleet, said he considered the following three tuning systems acceptable for powerplant use:

  • PSM’s auto-tune.
  • Wood Group GTS’s Ecomax™.
  • The OEM’s OpFlex™ AutoTune.

The controls expert said PSM’s auto-tune was installed on three GE engines in his fleet, Ecomax on two. OpFlex AutoTune was not considered for deployment because of its cost, he added.

Most of the speaker’s experience was with PSM’s offering, which, at the time of the meeting, had accumulated more than 8000 hours of operation integrated with Mark V and Mark VI control systems. Performance has been validated in all seasons and at ambient temperatures from 20F to 100F. Feedback from the plants has been good, he said.

Benefits of the system, he continued, were these:

  • Eliminates the need for seasonal manual tuning.
  • Reduces the risk of lean blowout (LBO) events.
  • Improves emissions control throughout the year.
  • Minimizes combustion dynamics, thereby maximizing hardware life.
  • Accommodates changes in gas properties while maintaining top performance.
  • Expectations are the same basic system will be suitable for Siemens gas turbines as well as GE’s. Development work on PSM auto-tune for the 501F is underway.

The PSM system, he continued, is customizable to meet the specific needs of each site. It is manufactured using industrial grade electronics, not Windows. Plus, it has an easy-to-use operator interface that is much like the Mark V screen.

Backgrounder. If you haven’t yet had the opportunity to dig into automated combustion tuning, given its rate of implementation, this may be a good time. The technology is complex, but the field of expertise is relatively small. At the top technology ladder for land-based gas-turbine CDMS applications is Dr Timothy Lieuwen, PE, assistant professor, Georgia Institute of Technology, School of Aerospace Engineering.

Equally comfortable in front of a class, in the lab, or on the deck plates of a powerplant, Tim, the name he prefers, is a breath of fresh air in an industry ripe for new ideas. Lieuwen has developed algorithms for predicting when LBOs are likely. Jim Fenton of Termecula (Calif)-based FocusTek said the algorithms basically count events that are precursors to LBO. These events increase in number as the LBO condition is approached and that information is used both to generate warnings and change combustion-process variables to avoid a complete loss of flame.

San Diego-based Alta Solutions Inc makes the CDMS “computer” that underpins automated tuning systems such as PSM’s. Lieuwen’s algorithms are incorporated into Alta’s offering for gas-turbine applications. The company’s AS-250 SpectralMon integrates advanced real-time monitoring and analysis with alarming criteria to detect changes in the dynamic signature of your engine. Once a dynamic event is detected, the system instantly informs the control system and records important data before and after the event. This information is used to change process settings and keep the engine “tuned.”

Others are involved in the field as well. Siemens Energy offers an automated tuning package, KEMA FlameBeat is in the mix, and EPRI is working with Lieuwen and others to analyze the various tones produced by the combustion process. “It’s all about the noise,” says Leonard C Angello, EPRI’s manager of combustion turbine technology.

Suggested reading at www.ccj-online.com to get you started: Click “Archives” on the tool bar at the top of the page and scroll to the issue indicated.

  • 3Q/2006, “Monitoring—and mitigating—combustion dynamics.”
  • 2Q/2008, “ 501F Users Group: Using advanced CDM analysis to improve reliability.”
  • 3Q/2008, “CDMS helps prevent forced outages, tune engine after overhaul.”
  • 3Q/2009, “ 7F Users Group: “Non-OEM F-class operational improvements gain traction.”

The editors followed up with a couple of plant managers to better understand how important automated combustor tuning might be to the reliable operation of their facilities. They said one of their primary concerns is the susceptibility of combustors to blowout of the lean flames required to maintain low NOx emissions and acknowledged the promise of automated tuning in preventing LBOs.

The discussion revealed that blowouts are embarrassing to operations personnel, reflecting an inability to maintain optimal combustion conditions. Undesirable consequences of a trip include (1) a lapse in power production, (2) wear and tear on the engine, which can shorten maintenance intervals, and (3) the emissions penalties associated with restarts.

The reaction of top management is yet another consideration. The financial types who occupy the executive suites of unregulated generation companies may not be aware of the damage combustion dynamics can cause their engines, but they certainly know when engine output drops to zero.

One of the plant managers who spoke to the editors said the gas turbines at his facility were prone to an occasional LBO while operators were getting used to the equipment during commissioning activities and shortly thereafter. However, since the units were tuned to rich PM3 with NOx at about 10 ppm, and an alarm was installed to warn if NOx drops to 7 ppm, there have been no incidents.

LBO generally is considered imminent by people knowledgeable on the subject when NOx drops below about 6 ppm. Come up to speed on engine tuning and LBO by reading “Preventing blowout trips of 7FAs,” which begins on p 16.

The plant manager said his engines are protected under an OEM long-term services agreement and are monitored by experts in GE’s diagnostic center. Under the terms of this LTSA, the cost of an LBO trip would not be charged against the plant. Engine tuning, as necessary, is done from the M&D Center.

Another plant manager said GE monitors dynamics at all frequencies as part of the LTSA. He pointed out that this is just good business practice because it protects hardware that the OEM is responsible for under the contract. The engines he manages behave differently—some well, some not so well. They have unique personalities, he said, and the plant has experienced a couple of LBO trips over the years. One unit tripped shortly after it was tuned. “All tuners are not created equal,” he added.

This facility pays for NOx tunes—that is, when the NOx level immediately downstream of the engine hits 7 ppm, a tune is ordered and conducted from the OEM’s M&D Center. The cost for this service is in the low five figures. The manager said this approach was most cost-effective for them.
It appears there are no easy answers with respect to tuning, which is a seasonal requirement for most engines as the ambient temperature goes from cold to hot and from hot to cold. Automatic tuning may make sense given the successes reported thus far—provided your control system’s capabilities are compatible with tuning system needs. However, there’s an expense associated with adding this capability.

Does every plant need automatic tuning? Certainly not. Your decision will be based on many factors—including level of combustion system sophistication, emissions permit requirements, control-system capabilities, dispatch contract, etc. One plant manager recalled hearing about a generation company with dozens of legacy gas turbines and favorable emissions permits that operated its engines for years at ambients between 20F and 100F without ever tuning.

Contrast that with LBOs attributed to the sudden deep freeze that hit the Southwest early in 2011and contributed to the havoc on the Texas grid. Had automatic tuning solutions been in place perhaps some of those trips might not have occurred.

Not all LBOs can be prevented by tuning. Consider the disturbance initiated by a 138-kV fault that remained on the transmission system for about 1.7 sec. During the protracted fault, local voltage went to near-zero, which effectively reduced area load and caused generators to accelerate. Six DLN-equipped gas turbines tripped as a result of LBO.

Here’s how that happened: As the machines accelerated in response to the frequency excursion, their compressors forced more air into the combustion chambers at the same time the governors reduced fuel input to reduce speed. More air and less fuel caused the flames in these engines to blow out and trip the units.

Onsite tuning is not cheap and sometimes it is inconvenient. Plants relying on manual tuning must get a technical expert to the plant as soon as operators see NOx emissions drifting downward into the zone of “concern.” The tuning may conflict with the affected unit’s dispatch schedule and arrangements for alternative generation may be required.

As for the time it takes to manually tune an engine, one expert suggested plants plan for a one-day outage. Rarely, however, would you expect tuning to extend beyond a 12-hr shift. The simplest tuning assignments are for simple-cycle peakers that run at base load when in service. The tuner’s primary concern is protecting against LBO at base load; combustion parameters can be forgiving on the ramp up and the ramp down. Such an assignment might take only four hours. Tuning a load-following peaker is more involved; dealing with a combined cycle having quaternary fuel and extended turndown is about the most complex assignment.

Turbine section

The moderator for the turbine session, one of the steering committee’s senior members, had one slide: An assembly drawing for a 7FA turbine. The exploded view showed first-, second-, and third-stage turbine wheels and buckets (92 in each stage), 1-2 and 2-3 spacers, bucket lockwire, etc.

For each of the parts—including the nozzles, buckets, shrouds, and honeycomb seals associated with each stage—the moderator asked if anyone in the room had issues or experiences to share. Questions were relayed to the group for ideas and solutions. Wise move: With more than 200 7F engine experts on location, someone was almost certain to know how to solve a colleague’s problem—and someone always did.

Keep in mind that the standard parts shown on the assembly drawing really aren’t “standard.” With 7221, 7231, and 7241 owner/operators represented at the meeting, and given the design variations within those model series, one or two people could be experiencing problems that no one else was. The group clinic approach to learning was a big success.

Safety practices, lessons learned

The steering committee had one voice when the subject of safety was introduced. The consensus view: The industry was doing a pretty good job regarding safety, but there certainly was more that could be done. Also, good communication is critical to a safe working environment.

One of the most interesting solutions presented involved making the turbine area safer for borescope inspections. Tie-offs, yo-yos, and conventional scaffolding are not quite right for inspection activities it was said. Some of the traditional options can be intrusive to the work being performed. Not the ideal.

A user presented on the design and installation of temporary access platforms that have served his company well. He showed a drawing of a six-section platform designed to cover the area between the engine split line and permanent grating. The aluminum sections are secured in place with turnbuckles. The platform system took two plant staffers about 80 hours to build. It can be installed for a borescope inspection in about two hours.

For details, access http://ge7ea.users-groups.com.

Power and lighting required for gas-turbine maintenance was covered in the final presentation on Tuesday. After reading the title of his presentation, the speaker asked rhetorically, “What gets written up on every safety walk down during an outage?” Next, he wanted to know how many in the audience had not had any safety hits on extension cords in water or running through doors.

It’s really simple, the presenter said, to safely provide for power and light. He showed the group how to line up electrical power when the gas turbine was offline, how to install wall penetrations to bring power outside, where to place purchased boxes with multiple GFCI outlets (over-current protection on each) for easy access by maintenance personnel, the value of light strings with magnets for flexible out-of-the-way positioning, etc.

Safety initiatives are a top priority at the nation’s generating plants powered by gas turbines. More than 35% of the entries received by CCJ for the journal’s Best Practices Awards program in 2011 were for safety. Two years ago (2009), there were about half as many safety entries as there were in 2011; in 2007, there were about half as many as in 2009. This year (2012), the editors expect the safety entries to top last year’s total.

Point is that if you’re looking to see what others are doing to make their plants safer, the CCJ archives is a good reference resource. Access www.ccj-online.com and click on the “Archives” button in the top tool bar. Then access the details on each entry in the first-quarter issues of each year.

Day Three: Generators
It’s no secret that the men and women who operate and maintain the nation’s gas-turbine-based peaking, cogeneration, and combined-cycle generating facilities focus most of their attention on the GT. There’s also the occasional holdover from the “age of fossil steam” to keep a sharp eye on the HRSG and steam turbine, if installed. But it’s the rare plant that has a true generator champion, especially in these times of the ever-shrinking O&M staff.

This fact was not lost on the 7F Users Group’s Steering Committee, which invited Mike Sendlak, manager of Electrical Engineering for Edison Mission Generation’s technical support team to conduct a tutorial on generators Wednesday morning.

There was considerable pent-up demand for practical information on generators that could be brought back to the plant and put to use. And Sendlak was just the person to handle the assignment. Although an electrical engineer by education and now in a staff position, his heart is in the plant and with the equipment.

That Sendlak’s presentation was toasted when his computer suffered a meltdown a few days before the meeting might not have been a bad thing for the users. He spoke without notes and in a casual manner, handling any question at any time and rapidly gaining the respect of an audience generally uncomfortable talking about generator technology.

Sendlak’s presentation is not something that can be summarized here in the traditional sense because it was a tutorial designed to help users feel more at ease with generators and impart useful information along the way. He took attendees in directions they wanted to go based on questions.
Sendlak started by introducing 7F users to the various components of generators and discussing the basics: what a given part does, the kind of care it requires and how frequently, how it can fail and what happens if it does, etc.

Then he spent another hour or so discussing the variables to monitor online, the tests to conduct offline, the data to collect, etc. Perhaps the most important takeaway from this portion of the presentation was the importance of “trending, trending, trending” data to gauge machine health, better plan maintenance outages, and avoid forced outages to the degree possible.

If you weren’t in Houston to hear Sendlak, you missed a valuable and constructive workshop. But most of the subjects he spoke to are covered in the comprehensive series of generator articles by Schenectady (NY)-based Consultant Clyde Maughan running in the CCJ (see p 40). Maughan is one of the world’s leading experts on generator O&M. His series focuses on generator monitoring, inspection, diagnosis, and root-cause failure analysis. You can review the first seven articles at www.ccj-online.com/maughan.

To keep up with generator technology, consider joining the International Generator Technical Community at www.generatortechnicalforum.org. No charge. It features a robust technical library and online forums where you can get your questions answered by experts. Again, no charge.

Day Three: Auxiliaries

The auxiliaries session immediately after lunch featured two user presentations. The first described what is thought to be the first conversion of the OEM’s gas-turbine control system to one offered by a third party; the second offered valuable experience with water-cooled liquid-fuel check valves on a dual-fuel 7FA. Details on the controls retrofit can be accessed at www.ccj-online.com, click “Archives” in the toolbar at the top of the page, click “2011 Outage Handbook” on the menu and scroll to “Industry first: Ovation replaces Mark V on 7FA.”

Water-cooled check valves. There has been significant discussion over the years at meetings of several user groups regarding the coking problem many owner/operators of dual-fuel engines experience with standard liquid-fuel check valves. After switching from oil to gas, the oil remaining in check valves, which are located close to the combustors, is exposed to high temperature.

Above about 250F, that relatively small amount of oil oxidizes. The resulting coke coats check-valve internal surfaces (and fuel lines as well) and restricts the movement of valve parts. Once this occurs, a check valve will not open and close properly until it is overhauled.

The most common trip during fuel transfer is on high exhaust-temperature spread—caused almost exclusively by check valves “hung-up” on coked fuel. Startups on oil when fuel-system components are fouled can be challenging as well—sometimes impossible.

A user discussed the check-valve challenges encountered at his plant, then spoke about the corrective alternatives considered and five years of experience with the solution selected.

Four alternatives were evaluated for mitigating the coking problem to improve reliability. Here are the results of that investigation:

  • Air cooling: Not possible because of space limitations.Lay-up the liquid fuel system after use and purge with nitrogen: Restricted operational flexibility. As much as a day might be required to prepare the fuel system for restart.
  • Convert to a recirculating or return-flow liquid-fuel system to keep the oil moving and prevent coking. Financially burdensome. Plus, the Mark V control system had no spare I/Os to accommodate that upgrade.
  • Change-out the conventional liquid-fuel check valves with water-cooled valves. The most cost-effective option. One valve required per can, a direct drop-in for the original. The only extra step was getting water for jacket cooling, but that was simple because a closed-loop cooling-water header was within reach.

The first and only issue encountered, inadequate cooling, was solved by removing a downstream restriction in the cooling-water discharge line. The water-cooled JASC (Tempe, Ariz) valves have been in service for about five years, the speaker said, and there have been no forced outages related to coking during that time.

Transfer and startup reliability with the OEM-supplied valves was less than 60%. Startup issues were almost all check-valve related; transfer reliability was affected by controls problems as well as coking. Today, the maintenance specialist continued, startup reliability is north of 95% and fuel-transfer reliability is more than 90%. CCJ